Zaki Brahmi

Energy Efficient Partitioning and Scheduling Approach for Scientific Workflows in the Cloud

Energy consumption is emerging as a new crucial issue of the Cloud Computing environments such as data centers. The problem of power consumption is more challenging especially in the context of scientific workflows deployment in the Cloud as they trigger intensive computational tasks and data manipulation steps which begets excessive data movement operations over communication networks. For instance, it was revealed that network devices consume up to one-third of the total energy consumption of Cloud data centers. In this paper, we propose an energy-aware approach for scientific workflows scheduling in the Cloud. In the first step, we propose a Workflow Partitioning for Energy Minimization (WPEM) algorithm that allows reducing the network energy consumption of the workflow and the total amount of data communication while achieving a high degree of parallelism. In the second step, we use the heuristic of Cat Swarm Optimization to schedule the generated partitions in order to minimize the workflows overall energy consumption and execution time. We evaluated the proposed approach using three real cases of data intensive workflows and compare it with other algorithms from literature. The experimental results show that our proposal allows to reduce remarkably the network energy consumption of the tested workflows (up to 96% of network energy consumption saving for memory intensive workflows) and the overall energy consumption of the workflows while ensuring a reasonable execution time and using less Cloud resources.

Enhancing Decentralized MAS-Based Framework for Composite Web Services Orchestration and Exception Handling by Means of Mobile Agents Technology

Decentralized orchestration offers performance improvements in terms of increased throughput and scalability and lower response time. However, decentralized orchestration also brings additional complexity to the system, mainly, in terms of exception handling. The research presented in this paper is carried out on the basis of some previous work of the authors, including: decentralizing orchestration of composite Web services and exception handling. We focus also on current works expanding the previous one, exhibiting thus a higher performance degree which the integration of mobile agents performs by moving the applications functionality through the network.

QoS-aware Automatic Web Service Composition based on Cooperative Agents

Automatic Web Service Composition (AWSC) is the processes of combining a chain of connected atomic services together in order to create a more complex and value-added composite service. To differentiate web services which have the same functionalities, Quality of Service (QoS) has been mostly applied. Given a high dynamicity and a rapid growth in the number of similar functionally web services, finding an efficient web service composition in a reasonable time satisfying a user requirements has become a challenging task. Many approaches in literature address the problem of QoS-aware automatic Web service composition. However, the majority of the existing approaches are restricted to predefined workflows and have limitations in terms of accuracy, scalability as well as dynamicity. In this paper, we propose a novel approach that solves major limitations encountered in the studied approaches. The proposed approach which is a set of cooperative autonomous agents is based on two mains ideas: 1) Self-organization of agents into dependency graph named social network agent, and 2) distributed computing of the optimal web services composition by a cooperative protocol among agents. Our approach can generate an accurate composition in a dynamic environment and is scale with the number of web services.

Semantic shared space-based complex tasks allocation method for massive MAS

Task allocation is still a fundamental problem in Multi-Agents System (MAS). It allows coalition formation of agents in order to cooperate together to perform a complex task. In general, the task allocation process includes two steps: i) finding the set of agents that can, potentially, participate to task allocation process, ii) computing the optimal allocation to execute the given task. In this work further attention is given for the first step. Indeed, in the context of massive MAS, characterized by dynamic, heterogeneous and a large number of autonomous agents, an efficient model of communication is required. This implies a need for a scalable and semantic infrastructure which allows: i) agents to be able to easily find each other and ii) semantic interoperability that refers to a common understanding of information communicated between agents. In this work information refers to an announced task. Different models of communication have been proposed, including broadcasting, forwarding, central server and group communication. Most of these approaches do not scale well in the context of massive MAS; when the number of agents grows. In additional, agent communication languages (ACLs), such as the KQML or FIPA ACL divide messages into several layers, and provide a specific syntax and semantics only for the outer layer, but its content is still arbitrary. To deal with these limitations, this paper extends our last task allocation method for massive MAS to shared space mechanism. This mechanism allows agents to find each by providing a logical shared space with temporal and special decoupling properties. To ensure semantic interoperability, we use a Task Ontology language (OWL-T) as a tuple space and a FIPA content message. OWL-T is based on the OWL for formally and semantically defining task in a high-level abstraction.

A Survey on Outlier Detection in the Context of Stream Mining: Review of Existing Approaches and Recommadations

Generally, extracting only expected knowledge from data is not sufficient since unexpected ones can hide useful information concerning the data behavior. These information can be further used to optimize the current state. This has lead to the outlier detection. It refers to the data mining task that aims to find abnormal points or sequence of data hidden in the dataset. In fact, due to the emergence of new technologies, applications often generate and consume data in form of streams. This data differs from the static one. Therefore, traditional techniques cannot be used. Hence, convenient ones suitable to the data stream nature must be applied. In this paper, we will review different techniques of outlier detection in the data streams. In addition, we shall describe different approaches based on these techniques in order to establish a comparative study based on different criterion. This study aims to help users and facilitates the choice of the appropriate algorithm for a certain context.

VM placement algorithm based on recruitment process within ant colonies

The Cloud computing forges the shape of the current era and the following ones based on delocalized IT infrastructure and sharing resources. However, the rebellious rise of cloud computing comes with concerns over energy consumption. Numerous reports which inspected Cloud energy consumption showed that the Cloud is an energy monster, specifically the data centers that holds 2% of overall energy consumed in the world on 2011 [6]. More closely, it has been proven that the servers (Physical machines PM) are the most energy-hungry elements of the data center [7]. Server consolidation based on virtualization is a key mechanism for energy consumption taming. Within this context, our aim in this paper is to propose a VM Placement Algorithm Based on Recruitment process within Ant Colonies proposed by Bonabeau [12] that seeks to maximize PM resources exploitation along with maximizing resources balance. The exprimental results showed that our algorithm generates always a significantly good solution.

Temporal Reconfiguration-Based Orchestration Engine in the Cloud Computing

In our days, the cloud computing wins a great importance. So it becomes the refuge of many companies especially Small and Medium sized enterprises (SMEs), since it provides computer services witch fits with demand and charged according to their use. Now the evolution towards the cloud is promoting that orchestration business process to be run as a service (Orchestration as a Service (OaaS)). OaaS represents a solution especially for (SMEs) which needs IT Systems intergration, but cannot install and use such integration platforms because of their maintenance costs and operation. OaaS is a specialization of paradigm Platform as a Service (PaaS). It reduces integration costs by outsourcing the operation and maintenance of an orchestration engine to an OaaS provider. The orchestration engine must be able to maintain its functionalities and performances in case of high demand. It has to be faster and the users have to pay less to run their orchestration processes. In this article, we propose an orchestration engine as a service based on the temporal reconfiguration approach. The proposed approach is based on two main ideas : i) Partition the amount resources of cloud server proportionally between BPEL processes. ii) Applying a temporal partitioning algorithm on a set of BPEL process. Our approach can be executed in a dynamic environment and is scaled with the number of BPEL processes.

Self-organizing Agents for Dynamic Network- and QoS-Aware Service Composition in Cloud Computing

Cloud computing plays a vital role in distributed systems used by Internet users. It provides a flexible environment in which data, equipment and services can be shared among end users in order to save time and cost. Cloud service composition is still one of the most important issues related to this paradigm. Indeed, Dynamic Cloud Service Composition (DCSC) is the process of combining a chain of connected atomic Cloud services together in order to create a more complex and value-added composite service. In this work, we present a new method of cloud service composition guided by QoS of services (execution time and cost) and network QoS (data transfer cost and latency). The latency is estimated by the Euclidean distance calculated based on the coordinates of Cloud services that are based on a network coordinate system called GNP (Global Network Positioning). The proposed solution is based on the paradigm of agents where autonomous entities cooperate together to generate an optimal composition within a reasonable time. Experimental results confirm the modesty of our approach.

Communication-aware VM consolidation based on formal concept analysis

In a Cloud Computing environment, a pool of resources in multiple physical machines is shared among virtual machines. Those virtual machines are deployed to host client applications and communicate together to run the appropriate tasks. Therefore communication between VMs should be taken into consideration when allocating VMs across servers. Recently, research works on network communication prove that communication between VMs should be seriously considered to save energy communication cost between network elements within a data center. Aiming to reduce resource wastage, energy consumption and energy communication cost, we propose in this paper a Grouping-based Virtual Machine Placement algorithm, based on Formal Concept Analysis (FCA), to allocate dependent VMs to servers as close as possible to each other. In order to demonstrate the effectiveness of G-VMP algorithm we compare the proposed with another algorithm. Experimental results show that G-VMP achieves better performance.

Cooperative agents based-decentralized and scalable complex task allocation approach pro massive multi-agents system

A major challenge in the field of Multi-Agent Systems is to enable autonomous agents to allocate tasks efficiently. In previous work, we have developed a decentralized and scalable method for complex task allocation for Massive Multi-Agent System (MMAS). The method was based on two steps: 1) hierarchical organization of agent groups using Formal Concepts Analysis approach (FCA) and 2) computing the optimal allocation. The second step distributes the tasks allocation process among all agent groups as follows: i. Each local allocator proposes a local allocation, then ii. The global allocator computes the global allocation by resolution of eventual conflict situations. Nevertheless, a major boundary of the method used to compute the global allocation is its centralized aspect. Moreover, conflicts process is a greedy solution. In fact, if a conflict is detected steps i) and ii) are reiterated until a non conflict situation is attained. This paper extends our last approach by distributing the global allocation process among all agents. It provides a solution based on cooperation among agents. This solution prohibits generation of conflicts. Its based on the idea that each agent picks out its own sub-task.